This portfolio summarizes projects I have been involved in to varying degrees, with a focus on LLM prompting and natural language processing (NLP). It includes examples from: academic and research-oriented work, freelance projects, my Data Science & AI bootcamp, and exploratory projects I pursue for fun.
I am currently preparing an extended and interactive digital version of this portfolio. In the meantime, I am happy to provide additional details or materials about any specific project upon request.
I am currently preparing an extended and interactive digital version of this portfolio. In the meantime, I am happy to provide additional details or materials about any specific project upon request.
| portfolio__cortés_rodríguez_Álvaro.pdf |
| projects__cortés_rodríguez_Álvaro.zip |
Overview of the projects
1. Guided multilingual translations using a modular prompt technique
I built a Jupyter Notebook script using OpenAI’s API to automate translations into multiple languages, ensuring they met customer requirements for style, tone, and cultural appropriateness. The translations were also aligned with the product’s supportive tone of voice.
I built a Jupyter Notebook script using OpenAI’s API to automate translations into multiple languages, ensuring they met customer requirements for style, tone, and cultural appropriateness. The translations were also aligned with the product’s supportive tone of voice.
2. Synapse: RAG-powered knowledge graph
For my final project in the Data Science & AI bootcamp, I helped develop a knowledge graph that mapped key concepts from the curriculum and enabled interactive exploration. We used retrieval-augmented generation (RAG) to provide dynamic, AI-powered guidance for aspiring data scientists.
For my final project in the Data Science & AI bootcamp, I helped develop a knowledge graph that mapped key concepts from the curriculum and enabled interactive exploration. We used retrieval-augmented generation (RAG) to provide dynamic, AI-powered guidance for aspiring data scientists.
3. BriefPoint: A news summarization app
I collaborated with a web developer to build an app that summarizes and organizes news from YouTube channels. It automatically generates concise, multilingual summaries and tags content by topic to make news more digestible and balanced.
I collaborated with a web developer to build an app that summarizes and organizes news from YouTube channels. It automatically generates concise, multilingual summaries and tags content by topic to make news more digestible and balanced.
4. Email classification
I developed an example showcasing how prompts can be used to automatically classify incoming emails for a SaaS project management company. The solution categorized emails into six business-relevant groups, assessed urgency, and generated short rationales to ensure customer-critical issues were prioritized over low-value noise.
I developed an example showcasing how prompts can be used to automatically classify incoming emails for a SaaS project management company. The solution categorized emails into six business-relevant groups, assessed urgency, and generated short rationales to ensure customer-critical issues were prioritized over low-value noise.
5. LLM usage in academic research
During my academic work in experimental linguistics and psycholinguistics, I integrated large language models (LLMs) to streamline repetitive tasks and explore complex linguistic phenomena. This early hands-on experience gave me a strong foundation in adapting LLMs for practical use, which I have since expanded through freelance and applied projects.
During my academic work in experimental linguistics and psycholinguistics, I integrated large language models (LLMs) to streamline repetitive tasks and explore complex linguistic phenomena. This early hands-on experience gave me a strong foundation in adapting LLMs for practical use, which I have since expanded through freelance and applied projects.
